Annals of Oncology Advance Access originally published online on July 28, 2007
Annals of Oncology 2007 18(10):1666-1672; doi:10.1093/annonc/mdm267
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
© 2007 European Society for Medical Oncology
gastrointestinal tumors |
Patient characteristics and stratification in medical treatment studies for metastatic colorectal cancer: A proposal for standardization of patient characteristic reporting and stratification
1 Department of Oncology, Haukeland University Hospital, Bergen, Norway
2 Klinik fur Innere Medizin, Oldenburg, Germany
3 Division of Biostatistics, Mayo Clinic Cancer Center, Mayo Clinic, Rochester, Minnesota, USA
4 Department of Oncology, Radiology and Clinical Immunology, Uppsala University, Uppsala, Sweden and Department of Oncology and Pathology, Karolinska Institutet, Stockholm, Sweden
* Correspondence to: Dr H. Sorbye, Department of Oncology, Haukeland University Hospital, 5021 Bergen, Norway. Tel: +44 1865 617058; Fax: +44 1865 617022, E-mail address: rowett.lewis{at}esmo.org
 |
Abstract |
|---|
 Top Abstract introductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Background: Prognostic factors have the potential to determine the survival of patients to a greater extent than current antineoplastic agents. Despite this knowledge, there is no consensus on, first, what patient characteristics to report and, second, what stratification factors to use in metastatic colorectal cancer trials.
Patients and methods: Seven leading oncology and medical journals were reviewed for phase II and III publications reporting on medical treatment of metastatic colorectal cancer patients during 2001–2005. One hundred and forty-three studies with 21 214 patients were identified. The reporting of patient characteristics and use of stratification was noted.
Results: Age, gender, performance status, metastases location, sites and adjuvant chemotherapy were often reported (99–63%). Laboratory values as alkaline phosphatase, lactate dehydrogenase and white blood cell count, repeatedly found to be of prognostic relevance, were rarely reported (5–9%). Stratification was used in all phase III trials; however, only study centre was used with any consistency.
Conclusion: There is considerable inconsistency in the reporting of patient characteristics and use of stratification factors in metastatic colorectal cancer trials. We propose a standardization of patient characteristics reporting and stratification factors. A common set of characteristics and strata will aid in trial reporting, interpretation and future meta-analyses.
Key words: colorectal neoplasm, metastatic disease, patient characteristics, prognosis, stratification
|
introduction |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Clinical trials, although in theory using similar patient selection criteria, often display surprising heterogeneity in response and survival rates. One likely explanation for this phenomenon is differences in patient characteristics or prognostic factors. Prognostic factors or their constellation have the potential to determine the survival of patients to a greater extent than any available antineoplastic agent or drug combination. Therefore, a critical component of a study report on treatment for cancer is the presentation of baseline patient characteristics. The patient characteristics table should convince the reader that the population included in the study, to the extent possible, is representative of the population with this type of cancer seen in clinical practice. This is particularly important in uncontrolled phase II studies, but is also relevant for phase III studies. Furthermore, the patient characteristics must show that there is no imbalance between the study arms if the study is randomized. However, although many of the known factors having prognostic value are reported, there is no consensus on what specific patient characteristics to report, and often important data are lacking.
Stratification factors are essential components of clinical trials. Appropriate stratification facilitates the simple interpretation of study results for primary endpoints in the presence of potentially confounding concomitant variables. Because it is not practical to stratify a randomization for all possible confounding influences, the challenge is to identify a minimal clinically relevant subset of variables to use as stratification factors. Inadequate use of stratification variables may result in chance imbalances between treatment arms, potentially causing a questioning of the conclusions of a study [1]. Performance status (PS) is one of the most useful stratification variables in oncology trials. It would be useful to supplement the prognostic ability of PS by a small number of factors to provide the best balance across treatment arms in regard to the proportion of patients who do well or do less well. There is no consensus on what to use as stratification factors in randomized metastatic colorectal cancer trials, although some recommendations have been made [2].
The joint influence of patient and tumour characteristics in patients with metastatic colorectal cancer on outcome is not well understood. To address this issue, Köhne et al. [3] performed a multivariate analysis to identify factors associated with survival, using a database of 3825 patients with metastatic colorectal cancer (mCRC), treated with 5-fluorouracil (5FU) and folinic acid, from 19 randomized studies. Eleven laboratory variables, seven tumour-related variables and five clinical variables were examined. The following factors were identified as having the greatest prognostic value: platelet count >400 x 109/l, alkaline phosphatase (ALP) >300 U/l, white blood cell count (WBC) >10 x 109/l, haemoglobin level <11g/l, PS (Eastern cooperative oncology group (ECOG) >1), number of tumour sites >1, and presence of liver or peritoneal metastasis. Median survival for an individual patient, predicted from these covariates, ranged from 4 to 16 months. Treatment with a more effective drug schedule may alter predictions based on the prognostic factors; however, the same factors also seem to be prognostic for survival after irinotecan and oxaliplatin combination treatments for mCRC [4–12]. The magnitude of effect from such prognostic factors can be profound. For example patients with normal ALP and platelet levels at baseline had a median survival of 23 months after first-line oxaliplatin-based chemotherapy, whereas median survival was 10 months if both levels were elevated [13]. Kabbinavar et al. [12] recently grouped mCRC patients receiving bevacizumab first-line combination treatment into high, intermediate and low risk groups using PS, number of tumour sites, ALP levels and WBC count, as suggested by Köhne. They found a median survival of 26 months in the low-risk group, 20 months in the intermediate-risk group and 14 months in the high-risk group.
As there is general knowledge of the importance of baseline patient characteristics for survival, we examined, first, how patient characteristics were reported in phase II and III studies of medical treatment for mCRC in the past 5 years in five leading oncology journals and two leading medical journals. From the identified publications, we pooled at a trial level the reported patient characteristics in an attempt to characterize typical patients included in mCRC medical treatment trials. Second, we identified what stratification factors have been used in the identified phase III studies.
|
patients and methods |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
Five leading oncology journals (Journal of Clinical Oncology, Annals of Oncology, British Journal of Cancer, Cancer, European Journal of Cancer) and two leading medical journals (The Lancet, New England Journal of Medicine) were reviewed for publications reporting the medical treatment of mCRC patients in phase II and III studies from 2001 to 2005. One hundred and forty-three studies were identified (Table 1). For each study, the reporting of patient characteristics was abstracted. In the 29 phase III studies, stratification factors were also examined.
|
study endpoints and statistical analyses
Statistical analyses were performed using the SPSS statistics package (v13.0, SPSS Inc. Chicago, USA). Median values of patient characteristics were calculated for the entire patient population, but also per study. When comparing differences in patient characteristics between first- and second-line studies and phase III versus phase II studies, the student t-test was used if the distribution was symmetric; if asymmetric, Mann–Whitney tests were used. Tests were considered statistically significant if P<0.05.
|
results |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
patient characteristics reported in the trials
One hundred and forty-three studies were identified, which included a total of 21 214 patients. The median number of patients in the studies was 69 (range 13–1120). The primary results of the reporting of patient characteristics reported are shown in Table 2.
|
age
One hundred and thirty-seven studies (96%) provided a median value, two gave the mean value. Seventeen studies reported the number of patients with older age, usually >70 years.
performance status
One hundred and twenty studies (82%) used the ECOG/WHO classification, of which most reported 0 vs. 1 vs. 2, but eight studies reported 0–1 vs. 2 and three studies 0 vs. 1–2. Eleven studies reported Karnofsky Performance Scale. Eight studies reported only mean values.
treatment of primary tumour
An indication of whether surgery of the primary tumour was performed was reported in only 26% of studies. In studies where this was reported, most specified that the surgery was removal of the primary tumour, others only that surgery was performed without any specification.
metastatic sites and location
The number of involved sites was not reported in one-third of studies. Most studies that presented data reported whether the number of sites was 1, 2 or >2; 27 studies reported 1 or >1 site. The median number of sites was given in four studies. The location of metastases was inconsistently reported, but included liver (n = 107), lung (n = 87), lymph nodes (n = 47) and peritoneum (n = 36). Reports of abdominal locations differed in description: abdominal mass, pelvic mass, recurrence, soft tissue, lymph nodes, intra-abdominal etc. Sixteen studies that reported exclusively on patients with only liver metastases/ peritoneal metastases were not included in the analyses of metastatic sites and location.
laboratory values
Very few laboratory values were reported, with no consistencies in characteristics reported. In general > Upper normal limit (UNL) values were used most often, but, many trials reported median values only. Nine studies reported the percentage of patients with elevated ALP values, seven studies the percentage of patients with elevated LDH values, and four studies the percentage of patients with WBC above 10 x 109/l.
others
Rarely (1–2 studies each) reported parameters included tumour area, body surface area, platelet count, albumin level, time from last chemotherapy, duration of metastatic disease, time from diagnosis of advanced disease, best response on first-line chemotherapy, bilirubin level, fever and locally advanced disease or metastatic disease.
patient characteristics of the entire study population
The characteristics of the 21 214 mCRC patients were calculated based on the median values provided from each study (Table 3). These data crudely reflect the characteristics of typical mCRC patients included in medical treatment trials. Phase II studies had a significantly higher percentage of patients with one metastatic site compared to phase III studies. Second-line studies had a significantly higher proportion of patients with PS 2, more metastatic sites, and a higher proportion of patients with lung metastasis. Patients with elevated ALP values ranged between 36 and 65%, with a median 50% (9 studies). The proportion of patients with elevated LDH values ranged between 31 and 69%, with a median of 42% (7 studies). Median time from initial diagnosis to treatment was 8 months (range 0–34 months). Median value for the proportion of patients with tumour-related symptoms was 61% (range 36–70%).
|
stratification factors
All 29 phase III studies used stratification factors. The median number used was 3, with a range of 1–7. The stratification factors used in these 29 trials are shown in Table 4. Four studies used centre stratification only. Centre stratification was in two studies by country, in one by clinician. PS was used as a stratification factor in 57% of the trials, typically stratified by ECOG/WHO 0 vs. 1–2. The number of sites was rarely used as a stratification factor; when it was, it was usually stratified by one versus greater than one site. Stratification factors used in one study only included bilirubin level, prior radiotherapy, prior pelvic irradiation, metastatic tumour size, percentage of liver involvement, status of disease, primary vs. recurrence of disease, prior immunotherapy, chemotherapy used and lung dominant vs. liver dominant disease. Previously used first-line chemotherapy was used as a stratification factor in three of the six second-line studies.
|
|
discussion |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
External validity, more commonly referred to as the generalizability of a trial, relates directly to whether the patients in the trial were similar to those patients in one's own practice for whom the results may be applied. Selection bias can seriously damage the external validity of a clinical trial by impairing the ability to compare whether the trial population in a given report is similar to that from other reports, or to the unselected clinical population [14,15]. Although randomized trials have the advantage of unbiased treatment comparisons (assuming appropriate trial conduct), they are not immune from selection bias. This bias can be introduced through the inclusion and exclusion criteria, but also in the patient enrolment process. Selection bias can be manifest not only through known prognostic variables such as age, income, PS and number of metastatic sites, but also for previously unknown or unstudied variables. For example patients who travel for elective care from specialized medical centres have better outcome than local patients with the same conditions who are treated at the same centres, and a longer travel distance from the treating centre has been associated with better survival [16]. It is a general phenomenon that participants in clinical trials tend to have better survival outcomes than non-participants [15]. Patients enrolled in clinical trials may bear little resemblance to the larger population of patients to which we wish to generalize the results because of the complicated processes by which patients are identified and recruited for clinical trials [17,18]. Adoption of less restrictive patient eligibility criteria has been suggested to increase trial generalizability [19]. Suggestions for improving the quality of reporting from medical studies have not focused on the problem of selection bias [20,21]. Martin et al. [22] recently found that the reporting of results in mCRC chemotherapy and surgical reports was limited to general outcomes, with a paucity of prognostic factors, which hinders the ability to compare results across treatments. Potentially, an objective and consistent, standard set of baseline criteria might allow corrected cross-trial comparisons, and facilitate meta-analysis.
patient characteristics
A problem for the generalizability of both phase II and III trials is the lack of reporting details of the study population and the larger target population [23]. In our study we found a vast inconsistency in the reporting of patient characteristics in published clinical trials. As there is a great variability among mCRC patients, the fact that important and independent prognostic factors are not presented at all or not presented accurately hinders the reader's ability to judge if a study could be biased in the selection of its patients. In our survey only age, gender and PS were regularly reported. Many of the factors known to impact survival, particularly laboratory values (see Table 5), were rarely reported. Newer studies have confirmed the importance of abnormal laboratory values as prognostic factors in mCRC, although there is little data on why these abnormal laboratory values evolve and what they reflect. Kabbinavar et al. [12] recently found that elevated ALP levels and WBC count had a major effect on survival after bevacizumab combination treatment. The CONFIRM 1 and 2 studies examined the effect of PTK/ZK, an oral antiangiogenetic compound, in first- and second-line treatment of mCRC patients. A pooled analysis of the two studies showed that PTK/ZK- treated patients with high LDH (>1.5 UNL) had significantly improved progression-free survival [30]. Preclinical and clinical data suggest that high serum LDH is a biomarker for highly angiogenetic tumours, and high LDH-5 in CRC is directly related to an upregulated HIF pathway and linked with an aggressive tumour phenotype [31]. Median or mean laboratory values have less relevance; rather, the proportion elevated or above a cut-off level is most informative. Unfortunately, even new, large studies present only mean laboratory values [28,32] or no values [33]. A non-conventional patient characteristics factor, but an important factor in mCRC, is whether the patient will be considered a potential surgical candidate if they respond to chemotherapy. If the secondary surgery rate exceeds the usual 10–15% rate, the definition of initial unresectability in the study has to be specified.
|
In describing the population of patients who entered these trials, compared with first-line studies, second-line studies had a significantly higher proportion of patients with PS 2, more metastatic sites, and more lung metastasis, probably reflecting a more advanced disease. We found surprisingly little difference between patient characteristics in phase II and III studies, as one might suspect more selected patients in phase II studies. However, the number of metastatic sites was higher in phase III studies, supporting such an assumption, but the general lack of reported important prognostic factors precludes any definite conclusion.
Based on our review, most studies do not provide adequate data concerning patient characteristics to conclude that the population included is representative of the patients with this type of cancer. There is therefore an urgent need for an improvement of mCRC study reporting of patient characteristics. Based on this review and Table 5, we have proposed a list of factors that as a minimum should be reported in future trials (Table 6). Age and gender are necessary in describing the study population. Concerning PS we suggest that only ECOG or WHO classification be used, as comparing these with the Karnofsky Performance Scale is difficult. Also, subdividing groups into PS 0/1 vs. 2 is more informative than 0 vs. 1/2. The optimal laboratory prognostic factors are unknown, but at this point ALP and LDH seem the best choices. ALP and PS have repeatedly been reported to be prognostic factors for survival (Table 5). PS and ALP at the start of first-line chemotherapy also predicts whether second-line chemotherapy will be given or not [34]. LDH, WBC and haemoglobin level are also among the important prognostic factors, but haemoglobin level might be altered before study inclusion with transfusions or erythropoietin. Site of primary and prior therapy are probably important patient characteristics although not always tested in the multivariate analyses for prognostic factors. More primary tumours are now left unresected and more patients receive adjuvant therapy. A minimum in reporting sites and location of metastases is to report 1 site vs. >1 site and liver vs. other, as these seem to have the greatest prognostic importance. Reporting mean/median values of PS, number of metastatic sites and laboratory values should be avoided. Race should be included in the report, as the use and development of germline polymorphisms will probably increase in clinical oncology.
|
Until further data evolve, Table 3, reporting pooled patient characteristics, could be used as a checkpoint for mCRC studies to examine whether the population in any particular trial differs in any substantial manner from the patients included in major studies the past 5 years. We acknowledge the limitations of the data from Table 3, which is based only on summary data at a study level. However, median values across studies calculated based on assuming the study-specific median value applied to all patients in the study were very similar to the median of the study specific values, supporting the robustness of the findings. Given the heterogeneity in study reporting, and the impractical nature of an individual patient data analysis, we consider Table 3 useful as a general guide.
stratification factors
A second fundamental issue related to baseline factors regards internal, as opposed to external, trial validity. Achieving balance on prognostic factors between treatment groups in a clinical trial is important to ensure that any observed treatment effect is appropriately attributed to the treatment itself. Improving balance on prognostic factors also potentially increases the statistical power attained in a trial. Substantial imbalances may occur by chance if simple randomization is used. Allocation of the treatment according to stratified random blocks, or dynamic allocation using a minimization routine based on clinical features are conventional approaches to obtain treatment groups that are as similar as possible. Stratifying at randomization on a factor is preferred to post hoc adjustment via Cox regression. Based on our study, the use of stratification factors in phase III studies varied. Centre stratification was used in 75% of the studies. Half the studies used PS as a stratification factor, and only four studies used laboratory values. In the studies the median number of factors used was 3, but varied from 1–7 factors. Kernan et al. have published guidelines for stratified randomization in clinical trials [35]. Stratification is recommended for small superiority trials with less than 400 patients. Even for larger trials, however, stratified randomization may be important when subgroup analyses or interim analyses are planned. For equivalence trials, Nam [36] suggests that randomization should be stratified even for larger sample sizes. The maximum desirable number of strata is unknown, but unnecessary stratification factors add complexity to a trial, and may reduce statistical efficiency. To suggest a standardization of stratification factors in mCRC trials is therefore difficult. We suggest that the following stratification factors should be considered as the most important: centre, PS, a laboratory value (ALP or LDH) and 1 metastatic site vs. >1. For second-line studies, stratification for prior chemotherapy or targeted therapy must be considered. Whether the patient will be considered a potential surgical candidate or not if they respond to chemotherapy can also be a valuable stratification factor in some trials. A common set of factors collected on all trials, with an allowance for one or two non-conventional trial-specific factors due to the individual needs of a trial, would clearly be a step forward.
In conclusion, we believe there is an urgent need for an international consensus on study reporting of patient characteristics and stratification in mCRC trials. Similar to the staging system for cancer, this consensus will clearly evolve over time, but until a baseline is established, what we are evolving from and to are unclear. A common set of characteristics and strata will aid in trial reporting, interpretation and future meta-analyses.
|
Acknowledgements |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
We thank Randi Eikeland (Clinical Cancer Research Office, Haukeland University Hospital, Bergen, Norway) for data management.
Received for publication January 12, 2007. Revision received April 3, 2007. Accepted for publication April 30, 2007.
|
References |
|---|
|
TopAbstractintroductionpatients and methodsresultsdiscussionAcknowledgementsReferences |
|---|
1. Bajorin D. The phase III candidate: Can we improve in the science of selection? J Clin Oncol (2004) 22:211–213.
2. Sloan JA, Loprinzi CL, Laurine JA, et al. A simple stratification factor prognostic for survival in advanced cancer: The Good/Bad/Uncertain index. J Clin Oncol (2001) 19:3539–3546.
3. Köhne CH, Cunningham D, Di Costanzo F, et al. Clinical determinants of survival in patients with 5-fluorouracil-based treatment for metastatic colorectal cancer: results of a multivariate analysis of 3825 patients. Ann Oncol (2002) 13:308–317.
4. Saltz LB, Cox JV, Blanke C, et al. Irinotecan plus fluorouracil and leucovorin for metastatic colorectal cancer. Irinotecan Study Group. N Engl J Med (2000) 343:905–914.
5. Mitry E, Douillard JY, van Cutsem E, et al. Predictive factors of survival in patients with advanced colorectal cancer: an individual data analysis of 602 patients included in irinotecan phase III trials. Ann Oncol (2004) 15:1013–1017.
6. Giachetti S, Perponit B, Zidani R, et al. Phase III multicenter randomized trial of oxaliplatin added to chronomodulated fluorouracil-leucovorin as first-line treatment of metastatic colorectal cancer. J Clin Oncol (2000) 18:136–147.
7. Freyer G, Rougier P, Bugat R, et al. Prognostic factors for tumour response, progression-free survival and toxicity in metastatic colorectal cancer patients given irinotecan (CPT-11) as second-line chemotherapy after 5FU failure. Br J Cancer (2000) 83:431–437.[CrossRef][Web of Science][Medline]
8. Bensmaine MA, Marty M, de Gramont A, et al. Factors predicting efficacy of oxaliplatin in combination with 5-fluorouracil (5-FU) ± folinic acid in a compassionate-use cohort of 481 5-FU resistant advanced colorectal cancer patients. Br J Cancer (2001) 85:509–517.[CrossRef][Web of Science][Medline]
9. Cunningham D, Pyrhonen S, James RD, et al. Randomized trial of irinotecan plus supportive care versus supportive care alone after fluorouracil failure for patients with metastatic colorectal cancer. Lancet (1998) 352:1413–1418.[CrossRef][Web of Science][Medline]
10. Rougier P, Van Cutsem E, Bajetta C, et al. Randomized trial of irinotecan versus fluorouracil by continuos infusion after fluorouracil failure in patients with metastatic colorectal cancer. Lancet (1998) 352:1407–1412.[CrossRef][Web of Science][Medline]
11. Pfeiffer P, Sorbye H, Ehrsson H, et al. Short-time infusion of oxaliplatin in combination with capecitabine (XELOX) as second line therapy in patients with advanced colorectal cancer after failure to irinotecan and 5-fluorouracil. Ann Oncol (2006) 17:252–258.
12. Kabbinavar F, Zurlo A, Irl C, Hurwitz H. Bevacizumab improves outcomes of patients with mCRC treated with IFL with or without bevacizumab independent of baseline risk. Proc Am Soc Clin Oncol (2006) 24:155s. abs. 3539.
13. Sorbye H, Glimelius B. Berglund Å et al. Multicentre phase II study of Nordic fluorouracil and folinic acid bolus schedule combined with oxaliplatin as first-line treatment of metastatic colorectal cancer. J Clin Oncol (2004) 22:31–38.
14. Zia MI, Siu LL, Pond GR, Chen EX. Comparison of outcomes of phase II studies and subsequent randomized control studies using identical chemotherapeutic regimens. J Clin Oncol (2005) 23:6982–6991.
15. Siu LL, Tannock IF. Problems in interpreting clinical trials. In: Handbook of Statistics in Clinical Oncology—Crowley J, ed. (2001) New York, NY: Marcel Dekker Inc. 473–490.
16. Lamont EB, Hayreh D, Pickett KE, et al. Is patient travel distance associated with survival on phase II clinical trials in oncology? J Natl Cancer Inst (2003) 95:1370–1375.
17. George SL. Selection bias, phase II trials and the FDA accelerated approval process. J Natl Cancer Inst (2003) 95:1351–1352.
18. Brawley OW. The study of accrual to clinical trials: can we learn from studying who enters our studies? J Clin Oncol (2004) 22:2039–2040.
19. George SL. Reducing patient eligibility criteria in cancer clinical trials. J Clin Oncol (1996) 14:1364–1370.
20. Begg C, Cho M, Eastwood S, et al. Improving the quality of reporting of randomized controlled trials: The CONSORT statement. JAMA (1996) 276:637–639.
21. Wils J, Sahmoud T, Sobrero A, et al. Evaluation of clinical efficacy of new medical treatments in advanced colorectal cancer. Results of a workshop organized by the EORTC GITCCG. Tumouri (1998) 84:335–347.
22. Martin RCG, Augenstein VA, Scoggins CR, et al. Quality of survival reporting in chemotherapy and surgical trials in patients with metastatic colorectal carcinoma. Cancer (2006) 106:1389–1394.[CrossRef][Web of Science][Medline]
23. Wright JR, Bouma S, Dayes I, et al. The importance of reporting patient recruitment details in phase III trials. J Clin Oncol (2006) 24:843–845.
24. Tournigand C, Andre T, Achille E, et al. FOLFIRI followed by FOLFOX6 or the reverse sequence in advanced colorectal cancer: a randomized GERCOR study. J Clin Oncol (2004) 22:229–237.
25. Fuchs CS, Moore MR. Harker G et al. Phase III comparison of two irinotecan dosing regimens in second-line therapy of metastatic colorectal cancer J Clin Oncol (2003) 21:807–814.
26. Wils J, Blijham GH, Wagener T, et al. High-dose 5-fluorouracil plus low dose methotrexate plus or minus low-dose PALA in advanced colorectal cancer in a randomised phase II-III trial of the EORTC Gastrointestinal Group. Eur J Cancer (2003) 39:346–352.[CrossRef][Web of Science][Medline]
27. Tournigand C, Cervantes A, Figer A, et al. OPTIMOX1. A randomized study of FOLFOX4 or FOLFOX/ with oxaliplatin in a stop-and-go fashion in advanced colorectal cancer – A GERCOR study. J Clin Oncol (2006) 24:394–400.
28. Kemeny NE, Niedzwieki D, Hollis DR, et al. Hepatic arterial infusion versus systemic therapy for hepatic metastases from colorectal cancer: a randomized trial of efficacy, quality of life, and molecular markers. J Clin Oncol (2006) 24:1395–1403.
29. Tampellini M, Saini A, Alabiso I, et al. The role of haemoglobin level in predicting the response to first-line chemotherapy in advanced colorectal cancer patients. Br J Cancer (2006) 95:13–20.[CrossRef][Web of Science][Medline]
30. Major P, Trarbach T, Lenz HJ, et al. Meta-analysis of two randomized, double-blind placebo-controlled, phase III studies in patients with metastatic colorectal cancer receiving FOLFOX4 and PTK/ZK to determine clinical benefit using progression-free survival in high LDH patients. Proc Am Soc Clin Oncol (2006) 24:153s. abs 3529.
31. Koukourakis MI, Giatromanolaki A, Simopoulos C, et al. Lactate dehydrogenase 5 (LDH5) relates to up-regulated hypoxia inducible factor pathway and metastasis in colorectal cancer. Clin Exp Metastasis (2005) 22:25–30.[CrossRef][Web of Science][Medline]
32. Kozloff M, Hainsworth J, Badarinath S, et al. Efficacy of bevacizumab plus chemotherapy as first-line treatment of patients with metastatic colorectal cancer: updated results from a large observational registry in the US (BRITE). Proc Am Soc Clin Oncol (2006) 24:155s. abs 3537.
33. Berry SR, Cunningham D, Michael M, et al. Preliminary safety of bevacizumab with first-line FOLFOX, XELOX. FOLFIRI and capecitabine for mCRC-first BEAT trial. Proc Am Soc Clin Oncol (2006) 24:154s. abs 3534.
34. Sorbye H, Berglund Å, Tveit KM, et al. Secondary treatment and predictive factors for second-line chemotherapy after first-line oxaliplatin-based therapy in metastatic colorectal cancer. Acta Oncol. 2007. in press.
35. Kernan WN, Viscoli CM, Makuch RW, et al. Stratified randomization for clinical trials. J Clin Epidemiol (1999) 52:19–26.[CrossRef][Web of Science][Medline]
36. Nam JM. Sample size determination in stratified trials to establish the equivalence of two treatments. Stat Med (1995) 14:2037–2049.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. Glimelius, M. Lahn, S. Gawande, A. Cleverly, C. Darstein, L. Musib, Y. Liu, K. L. Spindler, J.-E. Frodin, A. Berglund, et al. A window of opportunity phase II study of enzastaurin in chemonaive patients with asymptomatic metastatic colorectal cancer Ann. Onc., November 9, 2009; (2009) mdp521v1. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Chibaudel, C. Tournigand, P. Artru, T. Andre, A. Cervantes, A. Figer, G. Lledo, M. Flesch, M. Buyse, L. Mineur, et al. FOLFOX in patients with metastatic colorectal cancer and high alkaline phosphatase level: an exploratory cohort of the GERCOR OPTIMOX1 study Ann. Onc., August 1, 2009; 20(8): 1383 - 1386. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Bystrom, A. Berglund, U. Garske, H. Jacobsson, A. Sundin, P. Nygren, J.-E. Frodin, and B. Glimelius Early prediction of response to first-line chemotherapy by sequential [18F]-2-fluoro-2-deoxy-D-glucose positron emission tomography in patients with advanced colorectal cancer Ann. Onc., June 1, 2009; 20(6): 1057 - 1061. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Qvortrup, M. Yilmaz, D. Ogreid, A. Berglund, L. Balteskard, J. Ploen, T. Fokstuen, H. Starkhammar, H. Sorbye, K. Tveit, et al. Chronomodulated capecitabine in combination with short-time oxaliplatin: a Nordic phase II study of second-line therapy in patients with metastatic colorectal cancer after failure to irinotecan and 5-flourouracil Ann. Onc., June 1, 2008; 19(6): 1154 - 1159. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. J.A. Punt and M. Koopman Capecitabine and Irinotecan As First-Line Treatment of Advanced Colorectal Cancer J. Clin. Oncol., April 10, 2008; 26(11): 1907 - 1908. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||